The main objective of this project is to complete a first order sequence to reactivity algorithm for "standard mechanism" protein inhibitors of serine proteinases in general, and for turkey ovomucoid third domain in particular. Such an algorithm should allow for the prediction of the association equilibrium constant with several different enzymes solely from the amino acid sequence of the inhibitor. It should greatly aid in the design of physiologically and pharmacologically important inhibitors of great strength and/or high specificity. Experimentally, the project consists of preparing a set of all possible coded variants (one substitution at a time) at all the contact positions of turkey ovomucoid third domain. There are l2 such positions. One of them is unvaried Cys16, which is also structured and therefore will not be varied. The most crucial residue, Leu18 (the P1 residue) was already done in one laboratory. This - leaves 10 positions with 19 alternatives at each position, and thus 190 variants. These are to be augmented by several noncoded variants to be prepared either by enzymatic semisynthesis, or by total synthesis. The coded variants will be obtained by recombinant DNA technology. The transformed bacteria will be provided from Rutgers for the first two years; during the last three, they will be made at Purdue. Expression, purification and characterization will be at Purdue during the whole project. The non-coded variants will be obtained either by enzymatic semisynthesis, a technique developed at Purdue, or by total synthesis. For each variant, the equilibrium constants for association with six serine proteinases will be measured. In most cases, the equilibrium constant for reactive site peptide bond hydrolysis, the melting temperatures of virgin and modified forms, and the rate constants k(on) and k*(on) will also be measured. Heats and heat capacities of association and hydrogen ion release will be measured when experimentally feasible to obtain large comparative sets. Numerous samples will be sent to others for X-ray crystallography. Several inhibitors for human leukocyte elastase, for enteropeptidase (formerly enterokinase) and for subtilisin-like proprotein convertases, including human furin, will be designed, expressed, and tested. Some of these may be useful as drugs. Aside from their relevance to protein proteinase inhibitors, the data appear to be highly relevant to problems of biomolecular recognition and of protein stability.